Organizational Dynamics (2017) 46, 120—132

Available online at www.sciencedirect.com

ScienceDirect

jo urnal homepage: www.elsevier.com/locate/orgdyn

Applying the Job Demands-Resources model:

A ‘how to’ guide to measuring and tackling work

engagement and burnout

Wilmar B. Schaufeli

INTRODUCTION In this paper I make the case that the Job Demands

Resources (JD-R) model can be used as an integrative con-

ceptual framework for monitoring the workplace with the

There is a great practical need to assess psychosocial factors

aim to increase work engagement and prevent burnout. The

at work and improve employee well-being. Evidently, poor

JD-R model is particularly suited for this purpose because:

working conditions and burned-out employees are asso-

(1) it integrates a positive focus on work engagement with a

ciated with, for instance, sickness absence, occupational

negative focus on burn-out into a balanced and comprehen-

injuries and accidents, poor work performance, and reduced

sive approach; (2) it is has a broad scope, that allows to

productivity, whereas the opposite is true for good working

include all relevant job characteristics; (3) it if flexible, so

conditions and employee engagement. So ultimately, psy-

that it can be tailored to the needs of any organization; (4) it

chosocial factors and employee well-being translate into

acts as a common communication tool for all stakeholders. In

financial business outcomes. Therefore it is in the company’s

contrast, previous models focused almost exclusively on

enlightened self-interest to monitor psychosocial factors at

negative aspects of the job and included a limited, pre-

work and employee’s well-being on a regular basis, so that

defined set of job characteristics. Because of its compre-

timely and targeted measures can be taken to prevent

hensive, broad, flexible and communicative nature the JD-R

burnout and to increase work engagement. Apart from this

model not only enjoys great popularity among academic

intrinsic reason there is also an extrinsic reason for organi-

researchers, but it makes the model also quite suitable

zations to monitor the workplace, at least in Europe. Follow-

for practical use in organizations.

ing the EU framework Directive 1989/391/EEC on

The paper starts with a brief description of the JD-R

occupational safety and health, all EU member states have

model and then introduces the Energy Compass, an online

issued legislation on the prevention of psychosocial risks at

survey tool that is based on the JD-R model. Next, a case

the workplace. The two most relevant provisions of that

example is presented of an organizational development

directive are that: (1) employers should ensure that all

project that illustrates the practical use of the JD-R model

workers receive health surveillance that includes psychoso-

for increasing engagement and preventing burnout.

cial risks (Article 15) and; (2) employers are held responsible

for preventing ill-health at work, and must take appropriate

measures to make work healthier (Article 5). Hence, orga-

A BRIEF OVERVIEW OF THE JD-R MODEL

nizations have a legal obligation to monitor psychosocial risk

factors and improve employee’s health and well-being.

fi

Although this legal framework only applies to EU member- The JD-R model was introduced about fteen years ago

states, monitoring risks and improving health and well-being to understand burnout, a chronic state of work related

is paramount in other countries as well because of immanent psychological stress that is characterized by exhaustion

advantages, including financial and business outcomes. But (i.e., feeling emotionally drained and used up), mental

how to achieve that, and what has occupational health distancing (i.e., cynicism and lack of enthusiasm), and

fi ’ psychology to offer? reduced personal ef cacy (i.e., doubting about one s

http://dx.doi.org/10.1016/j.orgdyn.2017.04.008

0090-2616/© 2017 Elsevier Inc. All rights reserved.

The JD-R model: A ‘how to’ guide 121

competence and contribution at work). After some years the Toon Taris and I showed that twelve studies confirmed the

model was supplemented with work engagement, a positive, mediating role of burnout in the stress process and of

fulfilling psychological state that is characterized by vigor engagement in the motivational process, whereas in the

(i.e., high levels of energy and resilience), dedication (i.e., remaining four studies partial instead of full mediation

experiencing a sense of significance, pride and challenge) was found for either burnout or engagement. More recently,

and absorption (i.e., being fully concentrated and happily we reviewed eight longitudinal studies among workers of

engrossed in one’s work). various countries and found that the predictions of the JD-R

According to the JD-R model, every job includes demands model regarding the causal relationships between job

as well as resources. Demerouti, Bakker, Nachreiner and characteristics (i.e., job demands and job resources) and

Schaufeli (2001; p. 501) defined job demands as “aspects employee well-being (i.e., burnout and work engagement)

of the job that require sustained physical or mental effort were largely supported by five studies, partly supported by

and are therefore associated with certain physiological and two studies, whereas only one study failed to find any long-

psychological costs”. Roughly speaking these are the itudinal relationship.

‘bad things’ at work that drain energy, such as work over- Although it was acknowledged that job demands and job

load, conflicts with others, and future job insecurity. resources could interact in affecting burnout and work

In contrast, job resources are the ‘good things’ that are engagement — e.g., co-worker support might buffer the

defined as “aspects of the job that may do any of the negative effect of work overload on burnout — the evidence

following: (a) be functional in achieving work goals; (b) for such interactions is rather weak. This led Xanthopoulou,

reduce job demands and the associated physiological and Bakker, Demerouti and Schaufeli (2009; p. 236) to conclude

psychological costs; (c) stimulate personal growth and that “ . . . the current evidence on demand  resource

3

development” . Examples of job resources are support from interaction effects shows that even if significant, the prac-

others (which helps to achieve work goals), job control tical relevance of such interactions tends to be low”.

(which might reduce job demands), and performance feed- Tw o other extensions of the JD-R model are worth men-

back (which may enhance learning). tioning. First, personal resources have been included in the

In essence, the JD-R model integrates two basic psycho- JD-R model. According to Xanthopoulou and her colleagues

logical processes. (2009; p. 236), these are defined as “positive self-evalua-

First, a stress process, which is sparked by excessive job tions that are linked to resiliency and refer to individuals’

demands and lacking resources may — via burnout — lead to sense of their ability to control and impact upon their

negative outcomes such as sickness absence, poor perfor- environment successfully”. Examples of personal resources

mance, impeded workability, and low organizational com- are self-efficacy, optimism, and organization based self-

mitment. Essentially, when job demands (the ‘bad things’) esteem. Although personal resources can be integrated

are chronically high and are not compensated by job successfully in the JD-R model and their effects may be

resources (the ‘good things’), employee’s energy is progres- substantial, it is still unclear which place they should occupy

sively drained. This may finally result in a state of mental in the model. This seems to depend on the personal resource

exhaustion (‘burnout’), which, in its turn, may lead to under study, for instance, stable personality traits (e.g.,

negative outcomes for the individual (e.g., poor health) as optimism) are more likely to act as antecedents of job

well as for the organization (e.g., poor performance). Sec- demands and job resources, whereas malleable personal

ond, a motivational process, which is triggered by abundant characteristics (e.g., self-efficacy) could act as mediators

job resources and may — via work engagement — lead to between job characteristics and well-being.

positive outcomes such as organizational commitment, Second, engaging leadership was added to the JD-R

intention to stay, extra-role behavior, employee safety, model. Engaging leadership is related to transformational

and superior work performance. In fact, job resources leadership, for instance both include inspiring followers. Yet

(the ‘good things’) have inherent motivational quality; they they also differ because engaging leadership: (1) is firmly

spark employee’s energy and make them feel engaged, rooted in a well-developed psychological theory of motiva-

which, in turn leads to better outcomes. tion (Self-Determination Theory; SDT); (2) specifically

Please note, that from an intervention point of view both focuses on how to increase employee engagement; (3) not

high job demands and poor job resources contribute to only includes an individual but also a social, team dimension.

burnout, whereas only abundant job resources (and not More specifically, engaging leaders: (1) inspire their fol-

low job demands) contribute to work engagement. Hence, lowers (e.g., by enthusing them for their vision and plans);

by increasing resources, such as social support, job control (2) strengthen their followers (e.g., by granting them free-

and feedback, two birds are hit by one stone: burnout is dom and responsibility); and (3) connect their followers (e.

prevented and engagement is fostered. In contrast, reducing g., by encouraging teamwork and collaboration). By inspir-

demands, such as work overload, conflicts and job insecurity ing, strengthening and connecting leaders promote the ful-

would only affect burnout but not work engagement. The fillment of follower’s basic psychological needs for

reason is that in addition to being potentially stressful, job autonomy, competence and relatedness, respectively, which

demands may also be challenging to some point so that according to SDT are inherent in all humans. It appears that

lowering job demands would result in less challenging jobs engaging leadership has an indirect effect on preventing

and hence lower levels of work engagement. For instance, burnout and increasing engagement by reducing demands

having to meet a tight deadline may also stimulate perfor- and increasing job resources, respectively. For instance,

mance. inspiring leaders provide their followers with organizational

The empirical support for the JD-R model is abundant. For resources (e.g., by emphasizing alignment, value congru-

instance, in a review published in 2014 of the JD-R model, ence, trust, and justice) and minimize their organizational

122 W.B. Schaufeli

Health impairment process

+ + Negative Job demands Burnout outcomes _

Work Positive Job resources

+ engagement + outcomes

Motivational process

Figure 1 The Job Demands Resources Model

demands (e.g. by circumventing bureaucracy and ade- Standards Indicator Tool (HSE MSIT), all of which have been

quately managing organizational change). Furthermore, developed in the 1990s in northwestern Europe. Although all

strengthening leaders provide their followers with work questionnaires include various job demands and job

resources (e.g., job control, use of skills, task variety) resources, they are not based on the JD-R model (see

and development resources (e.g., performance feedback, Fig. 1), or of any other job stress model for that matter.

career perspective), and monitor their qualitative and quan- Rather than being deduced from an overarching, conceptual

titative job demands (e.g., work overload, emotional framework, they have been developed inductively so that

demands, and work-home interference). Finally, engaging the masses of data that are being produced by these surveys

leaders connect their followers by providing them social are difficult to interpret.

resources (e.g., good team atmosphere, role clarity). In other So far the Energy Compass (EC) is the first instrument that

words, engagedleadersreducedtheirfollowers’ job demands, is explicitly based on the JD-R model. Its brief and informa-

which in its turn, reduced their levels of burnout and they tive name expresses that the EC might guide individuals as

simultaneously increased their followers’ job resources, well as organizations in choosing the right direction to find

which in its turn, boosted their levels of work engagement. energy at work. In addition to being based on a validated

In conclusion, the JD-R model is a rather straightforward conceptual framework, the EC is also more efficient and

and empirically validated model that specifies relationships more balanced (see Table 1).

between job (and personal) characteristics, leadership, This means that compared to the other psychosocial

employee well-being, and outcomes. Basically, it states that questionnaires, the EC uses shorter scales and suffers less

decreasing job demands, increasing job (and personal) from negativity bias because it includes more positive con-

resources and stimulating ‘engaged’ leadership prevents structs.

burnout and increases work engagement. And as a result In a review of 9 studies on the JD-R model, Toon Taris and I

of this, less negative and more positive outcomes are (2014) identified 30 potential job demands, 31 job resources,

achieved for both employees and organizations. Since job 22 outcomes, and 12 personal resources, whereby some

demands and job resources spark the health impairment concepts overlapped to a large degree (e.g., team harmony

and motivational processes, respectively their proper assess- and team cohesion). Taking this overlap into account as well

ment is paramount. Therefore, a specific assessment tool

that is based on the JD-R model was developed and will be

Table 1 Characteristics of Psychosocial Questionnaires

discussed below.

# Items # Constructs Efficiency Negativity

bias

AN ONLINE JD-R ASSESSMENT TOOL: THE

1

ENERGY COMPASS QEEW 203 27 7.7 3.3

COPSOC 135 30 4.5 3.0

QPSNordic 145 30 4.8 2.5

Numerous comprehensive surveys exist to evaluate psycho-

HSE MSIT 35 7 5.0 2.3

social factors at work, for instance, the Questionnaire on the

EC 133 58 2.3 1.3

Experience and Evaluation of Work (QEEW), the Copenhagen

Psychosocial Questionnaire (COPSOC), the Nordic Question-

Note: QEEW = Questionnaire on the Experience and Evaluation of

naire for Psychosocial and Social Factors at Work (QPSNordic), Work; COPSOC = the Copenhagen Psychosocial Questionnaire;

and the Health and Safety Executive’s Management QPSNordic = the Nordic Questionnaire for Psychosocial and Social

Factors at Work; HSE MSIT = Health and Safety Executive’s Man-

agement Standards Indicator Tool; EC = Energy Compass; Effi-

ciency = #items/#constructs; Negativity bias = #negative

1

This is a proprietary measure, the copyright is owned by Triple i

constructs/# positive constructs.

Human Capital, The Netherlands.

The JD-R model: A ‘how to’ guide 123

Table 2 Content of the Energy Compass

as my personal experience regarding the practical relevance The validity of the EC was investigated by testing the

of these variables for organizations, 12 job demands, 22 job underlying JD-R model. First a factor-analysis of job

resources, 9 outcomes and 8 personal resources were demands and job resources was carried out from which three

selected (see Table 2). Three job demands were included types of demands emerged (i.e., qualitative, quantitative,

that were deemed important from a practitioner’s point of and organizational demands) and four types of resources (i.

view but did not appear on the review (i.e., under-load, e., social, work, organizational, and development

bureaucracy, and pace of change), as well as four job resources); see Table 2. Next, using a two-stage approach,

resources (i.e., team effectiveness, fulfillment of expecta- the factor-scores of demands and resources were used as

tions of others, being aligned with the organization, and indicators of the corresponding higher-order constructs.

value congruence between employee and organization), and These constructs were included together with burnout, work

three personal resources (i.e., setting one’s own limits, goal engagement, outcomes and leadership into a so-called struc-

directedness, and self-development). The technical details tural equation model. In fact, this model can be seen as an

of the construction of EC are explained in Appendix A. operationalization of the JD-R model. It appeared that the

Immediately after completing the EC respondents receive hypothesized structural equation model fitted to the data of

an automatically generated feedback report in which their the representative sample. In other words, the relationships

scores are compared with those of the average employee. between the constructs that are included in the EC agree

The structure of the report, which can be saved or printed in with the propositions of the JD-R model (see Fig. 1).

pdf-format, follows the logic of the JD-R model with sepa- In conclusion, the EC consists of a broad set of valid and

rate sections for job demands, job resources, personal reliable indicators of job demands, job resources, outcomes,

resources, well-being, and outcomes. In addition, specific and personal resources that have been identified by previous

suggestions are given in this personalized feedback report research and consultancy experience. It is an efficient tool

about how to decrease high demands and increase poor that assesses a broad variety of constructs with relatively

resources. few items, thereby balancing a positive and a negative

124 W.B. Schaufeli

approach. Reference data are available that allow the clas- Step 1: Aim and Project Team

sification of employee’s scores as ‘high’, ‘average’, or ‘low’

as compared to the working population. So far, only Dutch

In the case of the hospitality project the main aim of top-

reference data are available (see Appendix A); for use

management was: how can employee’s levels of work

in other countries local reference data are recommended.

engagement be increased? Since according to the JD-R model

And last but not least, it was shown that the JD-R model

(Fig. 1) burnout can be reduced simultaneously, a secondary

serves as an overarching, conceptual framework for the

aim was added: to decrease levels of burnout. So the overall

EC. Or put differently, that the EC can be used to assess

aim of the project was to increase work engagement and

the JD-R model.

prevent burnout. The commitment of the organization to the

project was illustrated by the fact that together with a

dedicated consultant, a member of the top-management

USING THE JD-R MODEL IN ORGANIZATIONS:

(i.e., the CFO) and the head of the HR-department consti-

A CASE EXAMPLE

tuted project team. This also meant that, if necessary, the

process could be speeded up because communication lines

Below a project is discussed that uses the JD-R model to

were short and extra resources (i.e., finances and time)

improve work engagement and prevent burnout in the hos-

could be readily made available.

pitality industry. Salanova, Agut, and Peiró (2005) found that

The project team drafted a detailed planning that

work engagement is particularly important in the hospitality

included various milestones, such as delivering reports to

industry because it is related to service quality and customer

the management of the participating hotels. Tw o months

loyalty. On the other hand, research carried out by Pienaar

were planned for the preparation phase (steps 1—3), one

and Willemse (2008) shows that workers in the hospitality

month for the fieldwork (step 4), two months for data-

industry run the risk of burning out because of demanding

analyses and reporting (step 5), one month for survey feed-

customers and long and anti-social working hours.

back (step 6), and from the 7th month onwards interventions

The role of the JD-R model goes beyond the fact that it

would take place (step 7). Finally, a one-year follow-up

lies at the core of the EC, which is used to analyze the

measurement with the EC was planned.

complexity of the organization’s reality and to tease out the

most relevant factors for increasing work engagement and

Step 2: Customizing the Energy Compass

preventing burnout. Perhaps even more importantly, the JD-

R model plays a crucial role in the communication with all

The EC was tailored to the needs of the organization by the

stakeholders that are involved in the project; employees,

project team. Key persons, such as HR officers, hotel man-

top management, middle management, line management,

agement, supervisors, and work council members were con-

HR-officers, and workers council. The model constitutes a

sulted to identify the most relevant job stressors, personal

‘common language’ and acts as a lens through which all

and job resources, and outcomes that should be included in

stakeholders see the organization and their personal and

the EC. Since it was important to reduce the length of the

professional situation.

survey, four less relevant job demands (mental demands,

The project was carried out in one of the largest Dutch

negative changes, work underload, and harassment) and

hotel chains that operates 22 hotels in the Netherlands and

three job resources (fulfillment of expectation, participa-

employs 1055 employees. So far, the organization had car-

tion in decision making, and alignment) were eliminated

ried out in-house, annual, mandatory employee satisfaction

from the EC. Moreover, only two personal resources (proac-

surveys. Although the response was usually high top manage-

tivity and goal-directedness) and three outcomes (team- and

ment was dissatisfied with for two reasons. First, the quality

organizational commitment and turnover intention) were

of the in-house survey was questioned because it lacked a

included because these were deemed most relevant. Rather

solid scientific foundation. Second, it was suspected that the

than focusing on increasing personal resources, the organi-

results of the survey were biased because they were linked

zation opted for increasing job resources. And since the main

via specific targets to hotel manager’s financial bonuses. In

aim was increasing work engagement, only commitment was

other words, a perverse incentive could not be excluded.

included as on outcome because it is more closely linked with

Hence, an independent party was invited to carry out a 2

engagement than either employability or job performance.

voluntary, state-of-the-art, scientifically based engagement

The version of the EC that was used in the project included

survey and would suggest improvements to increase work

34 constructs (see Table 1).

engagement, as the organization aspired to employ

‘engaged’ rather than merely ‘satisfied’ employees. It was

Step 3: Internal Communication Campaign

reasoned by top-management that an engaged workforce

would realize the organization’s core values of respect,

Before carrying out the survey an internal communication

entrepreneurship, passion, style, and customer orientation

campaign was launched. Its goal was to stimulate employees

better than a satisfied workforce. In fact that is a realistic

to fill-out the EC and to be actively involved in the project.

expectation, which is by and large backed up by research and

The communications not only stressed the importance for

case studies.

the organization, but also the voluntary nature of the survey,

Fig. 2 shows the general process model — also called

and the fact that participating employees would receive a

regulative cycle — that is employed in organizational change

projects with the JD-R model and which consists of eight

steps. Each step will be discussed below and illustrated by

2

the hospitality project. Schaufeli (2014).

The JD-R model: A ‘how to’ guide 125

Figure 2 The Process Model of the Project

personalized feedback report of their scores on the EC. addresses, private email addresses were deliberately used

Please note that this was in sharp contrast to the previous in order to stress that participation was voluntary. During the

surveys that were mandatory and did not produce immedi- three-week period that the web link was open, hotel man-

ate, personalized feedback. Anonymity was guaranteed, and agers received five updates of the response rates of their

it was explicitly stated that nobody in the organization would own hotel, as compared to the other hotels. This was meant

have access to the data of the employees. Last but not least, to facilitate participation, as low scoring hotels were

it was emphasized that the results of the survey would be expected to stimulate their employees to fill out the EC.

used to improve employee’s working conditions. So the The final response rate was 43% (N = 452) and ranged

campaign provided two answers to the question ‘what’s in between hotels from 21 to 73%. Probably, this relatively

it for me?’: a personalized report as well as better working low response rate should be interpreted as a counter-reac-

conditions. tion to previous surveys that have been mandatory for all

The company’s intranet was used for informing all employees. The lowest response — about 20% — was

employees about the aims and planning of the project. observed among dishwashers and apprentices, most likely

For instance, a short 1.5-min animation was published because their commitment to the organization is relatively

on the intranet and posters were on display in all participat- low, also given high turnover rates for these groups. The

ing hotels. To stress the importance of the project, top- largest professional groups in the sample were desk recep-

management was the sender of all communications to the tionists (38%), waiters (22%), cooks & chefs (11%), and house-

employees. keeping (9%). The mean age of the sample was 36.6 years,

with 49% aged under 35 years, and an average tenure of

5.5 years. The figures illustrate the young age of the com-

Step 4: Survey and Individual Feedback pany’s workforce. The majority of the sample were women

(55%) and on the average employees worked 35.4 h per

All 1055 employees received an email in their private mail- week, which was 15% more than is stipulated in their labor

box with a link to the online EC. Instead of work email contracts. So overwork is pervasive.

126 W.B. Schaufeli

It took participants on average 42 min to complete the resources that should be improved in order to increase work

EC, varying from 12 min to over 3 h. Obviously some employ- engagement, a so-called priority analysis was carried out for

ees left the EC-window open on their computer while doing engagement (see Fig. 3a) and burnout (see Fig. 3b) sepa-

other things. The vast majority of the employees — 86% — rately.

needed maximum 30 min to complete the EC, which was set Basically, this analysis combines the level of employee’s

as the targeted time by the project team. job demands and job resources (above or below the bench-

Immediately after completing the EC, the employees mark) with the impact these scores have on work engage-

received an automatically generated feedback report, which ment and burnout (increase or decrease). Specifically, those

compared the employee’s scores with the benchmark (i.e. demands and resources that deviate from the benchmark —

the average Dutch employee) and which explains the mean- either positively or negatively — are entered in a regression

ing of the scores in greater detail. In case of an unfavorable analyses in order to determine their positive or negative

score relative to the benchmark the feedback text invites impact on work engagement and burnout.

the employee to take action. For instance, if a score indi- From the priority analyses of work engagement no job

cates that career perspectives are poor, a web link to the characteristics emerged that should be tackled or moni-

company’s career counseling service is provided for making tored. However, recognition (by customers and colleagues),

an appointment. The individual feedback report is struc- congruence of personal and organizational values, and use of

tured according to the JD-R model, meaning that job skills were identified as assets that increase levels of engage-

demands, job and personal resources, well-being (work ment and should therefore be utilized. In other words, these

engagement and burnout), and outcomes are each presented three resources are crucial in maintaining high levels of

under different headings. In order to assist employees with engagement. Most importantly, however, person-job fit,

the interpretation their report, the logic of the JD-R model is possibilities for learning and development, availability of

briefly explained. tools, and team-effectiveness scored below the benchmark

and should thus be further developed in order to enhance

levels of engagement.

Step 5: Analyses and Reporting

A similar priority-analysis for burnout revealed that skill

utilization should be utilized because this is an asset that

A general report was drafted for the entire organization, as

decreases burnout. Furthermore, there is room for improv-

well as for each of the 22 hotels individually. In addition,

ing person-job fit that should be developed in order to

reports were also drafted for the 14 largest teams that

decrease levels of burnout. Finally, physical demands, inter-

included more than 7 participating employees. The com-

personal conflicts, and emotional demands scored below the

pany-, hotel- and team-reports were based on aggregated

benchmark–— which is fine, of course. But because these

data, which means that average scores for the whole orga-

demands may potentially increase levels of burnout they

nization, the hotel, and the team are reported. Like the

should be monitored.

individual feedback report, the aggregated reports give an

In sum, although levels of engagement are relatively high,

overview of the scores for each element of the JD-R model,

it seems that further improvement is possible, specifically by

including a comparison with the benchmark. For the orga-

developing a limited set number of resources (person-job fit,

nization as a whole the average Dutch employee served as a

possibilities for learning and development, availability of

benchmark, whereas for the hotels and the teams other

tools, and team-effectively). Likewise, burnout could be

hotels and other employees from the same hotel were used

further reduced by increasing person-job fit and the utiliza-

as reference groups, respectively. For the sake of brevity,

tion of employee’s skills. It seems that there is a firm basis for

only the overall findings for the entire organization will be

this kind of improvements as results of the EC indicate that

discussed here. Needless to say that the results of individual

employees feel committed to the organization and their

hotels (and teams) may deviate from this general picture.

team, experience a supportive, positive and fair social

It appears that, generally speaking, scores on the EC are

climate, share values with the organization, and have trust

quite favorable. For instance, 28% of the employees were

in leadership.

engaged, whereas only 8% reported burnout complaints,

against 15% for both in the working population. Also levels

of job satisfaction and commitment were higher, whilst Step 6: Survey Feedback

turnover-intention was somewhat lower. On the negative

side, two stressors stand out; work overload was more The general report was discussed with top-management,

prevalent and the pace of change was too high for 5% of after which all 22 hotels received separate reports, as well

the employees, against only 1% in the reference group. As far as the 14 teams. Feedback sessions were held individually

as job resources are concerned, appropriate tools were with the management of the four most problematic hotels,

available for 49% of employees (60% in the reference group), whereas results were discussed with the remaining 18 hotels

and for 52% their pay was fair (66% in the reference group). In in four sessions with four hotels each. In addition to the

addition, scores on person-job fit, possibilities for learning dedicated consultant and the head of HR, the COO partici-

and development, and team effectiveness were (somewhat) pated in each feedback session as to emphasize the impor-

lower as well. Nevertheless, taken together, it seems that tance of the sessions for the top-management. In these

employees are exposed to few stressful job demands and can sessions two questions were answered.

draw upon considerable job and personal resources. First, in how far does the pattern of results correspond

This is not to say that job characteristics cannot or should with the perception of the stakeholders? This is a kind of

not improve. In order to identify those job demands and job validity check. In case the results of the EC would have been

The JD-R model: A ‘how to’ guide 127

(a) Above benchmark

Utilize

• Recognition

Tackl e • Value congruence

• Use of skills

Dec rease Increase

Deve lop

• Person-job fit

• Learning possibilities

Monito r

• Availability of tools

• Team effectiveness

Below benchmark

(b) Above benchmark

Utilize

Tack le • Use of skills

Increase Decrease

Monitor

• Physical demands Develop

• Emotional demands • Person-job fit

• Conflicts

Bel ow benchmark

Figure 3 (a) The Priority Matrix for Engagement. (b) The Priority Matrix for Burnout

completely at odds with the picture the stakeholders have, tailor-made procedure was followed here, whereby

something is wrong and the cause of that discrepancy needs the entire organization and each hotel (and team) formulated

to be identified before proceeding further. However, as a its own objectives. The reason for this was that — as noted

rule, the stakeholders shared the conclusions of the EC above — the results of the EC varied between various hotels.

and it was felt that the pattern of results that was uncovered Below in the next step the measures are discussed that

by the EC matched the perception of the general- and hotel were taken at the general, organizational level.

management. So it seemed that the EC produced a valid The feedback sessions were conducive in building a com-

picture. mon language to discuss the results and the actions to be

Second, given the conclusions of the EC, what measures taken. The terminology and the logic of the JD-R model,

could and should be taken in order to increase engagement which is quite easy to understand also for non-specialists and

and prevent burnout? Evidently, answering this question non-academics, proved to be quite helpful in this respect. In

includes prioritizing various potential measures. A addition, feeding back the results and discussing these

128 W.B. Schaufeli

critically with management, supervisors, and employees is table discussions at lunchtime with all employees. This

crucially important to build commitment and trust for imple- way, top-down and bottom-up communication is im-

menting interventions. proved, so that top-management is better informed

about the issues of employees at the shop floor, whereas

employees receive first-hand information about the or-

Step 7: Interventions

ganization’s policies. As a result, these site visits and

In principle, based on the results of the EC, two types of round tables might increase employee’s perceived rec-

measures can be taken. First, employees themselves may ognition and their shared values with the organization.

take measures to improve their own personal or job 4. Additional channels were opened to communicate with

resources, or decrease their job demands. Usually about employees, such as mailings to their private addresses

—

10% 15% of the employees do so spontaneously; for (both via email and surface mail), organize town hall

instance, they talk to their bosses or their colleagues to meetings with employee in hotels, and displaying posters

address certain issues, contact a career counseling service, in staffrooms with important messages. The aim of these

or consult their occupational physician. In the current pro- actions was to increase the alignment of the employees

ject, the focus was not on individual-based measures but on with the company’s mission and core values (i.e., re-

measures that were taken at the level of the entire organi- spect, entrepreneurship, passion, style, and customer

zation or the hotels or the team. For reasons of economy I orientation). This was expected to improve communica-

will only discuss the main measures that were taken at the tion and value congruence with the organization.

overall organizational level. In addition, each hotel or team 5. ICT-systems were updated and better adapted to the

fi

initiated speci c actions to improve engagement and pre- business process and to the user’s needs. Many employ-

vent burnout. For instance, in one particular hotel a large ees use these tools, and it was expected that they would

fi

gap existed between the front of ce (reception) and the benefit from improved ICT-systems when performing

restaurant, which had a negative impact on customer ser- their jobs.

vice. It was decided that receptionists would occasionally

work in the restaurant, and vice versa, in order to foster Step 8: Evaluation

mutual understanding. Furthermore, particularly for hotels

in Amsterdam the combination of a very high workload and a One year after the EC was carried out a follow-up measure-

fi

erce competition on the hospitality labor market resulted ment was conducted using the same survey. Again, and

in a shortage of personnel. Therefore, Amsterdam hotels following the same procedure, all employees were invited

decided to focus their HR policy more on personal retention. to fill-out the EC and 475 employees did so. However, the

Based on the report that was discussed with top-manage- records of only 241 employees could be linked as they

ment the following actions were taken for the entire orga- participated both times in the survey. This corresponds to

nization: roughly half (53%) of the initial sample. It appeared that in

this group the rate of engaged employees had increased by

1. The annual HR-cycle was updated with the aim of keep- 2% at follow-up, whereas the level of burnout remained the

ing jobs challenging. Traditionally, each summer employ- same. In the total follow-up sample, which also includes

ees had a performance review with their supervisor that those who filled in the EC for the first time, the rates for work

focused almost exclusively on past performance. This engagement and burnout were 32% and 8%, respectively. This

review was followed in the winter by a formal perfor- corresponds to an increase of 4% for engagement and a

mance appraisal. The performance review was replaced decrease of 1% for burnout. So, on balance, the new group

‘ ’

by a feed forward session , which is future oriented and who participated in the survey for the first time at follow-up

’

focuses on employee s preferences, potentials, and was more engaged and had less burnout complaints than

strengths, and how to develop these. In addition, a those who also participated one year ago. Moreover, an

‘ ’

talent chart is introduced which gives an overview of increase in job satisfaction (4%), organizational commitment

the talent pool at the level of the entire organization. (4%), and team commitment (3%) was observed, as well as

This information is used for HR policies to support career and a minor decrease in turnover intention of 2%. Hence, it

management of employees. Refocusing the HR-cycle seems that, overall, employee well-being, satisfaction, and

fi fi

would speci cally increase person-job t and the use commitment have increased. Also, it seems that those who

of skills, and therefore most likely enhance team effec- filled out the questionnaire for the first time at follow-up (i.

tiveness. e., new hires) are more engaged, satisfied and committed

’

2. The curriculum and usage of the organization s own than those with more tenure.

fi

training center was adapted to better meet the speci c Thus far it’s mostly good news. However, it is rather

needs of the employees. Their training needs were challenging to explain these positive changes since most

systematically inventoried, new training programs were drivers of well-being did not change over time. Rather than

launched, and existing programs were updated. The new change, stability seems to be the norm in the hotel chain.

range of training programs was communicated to hotel Nevertheless, some minor negative as well as positive

managers and their employees. Tw o internal trainers changes were observed. On the negative side, pace of

were dedicated to this project that, in essence, focused change, work-home conflict, and interpersonal conflicts

’

on increasing employee s possibilities for learning and slightly increased, whereas use of skills and value congru-

development. ence somewhat decreased. On the positive side, recogni-

fi

3. Top-management, personi ed by the CEO and the COO, tion, available tools, fair pay, organizational justice,

performed site visits to each of the hotels to have round possibilities for learning and development, and career

The JD-R model: A ‘how to’ guide 129

perspective have increased. The increase of some of these development process that aims to increase work engage-

job resources might be linked to specific measures that have ment and prevent burnout. The model is well equipped for

been taken. More specifically: (1) renewing the annual HR- this purpose because it is comprehensive, as it includes both

cycle might have increased employee’s recognition and a positive motivational process as well as a negative a stress

career perspectives; (2) renewing the training curriculum process. This balanced approach is an important asset to

might have increased employee’s possibilities for learning ‘sell’ the model to organizations because it integrates an

and development as well as their career perspectives; (3) occupational health approach (reducing job stress and burn-

site-visits and round tables, might have fostered employee’s out) with an HR-approach (increasing work motivation and

recognition by management, and organizational justice; (4) engagement).

opening new communication channels might have fostered Moreover, the JD-R model can be broadly applied in

recognition and organizational justice as well; and (5) updat- various types of organizations because a wide range of

ing and adapting ICT-systems might have improved the avail- job- and personal characteristics, as well as outcomes can

ability of tools. Yet, other resources remained unaffected or be included. At the same time the model is also flexible, in

even worsened, although not substantially. the sense that it may be tailored to the specific setting in

Taken together, the results of the evaluation are some- which it is applied. This is exemplified by the EC, which is an

what difficult to interpret, particularly as far as the effects efficient online tool to assess all relevant JD-R constructs.

of the measures are concerned. It should be kept in mind Potentially about sixty constructs can be included in the EC,

though, that only measures for the entire organization were but usually fewer are included as it is tailored to the orga-

taken into account here. As pointed out above these were nization. For instance, in the hospitality project 34 con-

supplemented by specific measures in particular hotels or structs were included, which took employees approximately

teams. Clearly, this complicates interpretation. Yet, the 30 min to complete. This combination of breadth and spe-

main message to be taken from the evaluation is three-fold: cificity is of great practical significance and is therefore

(1) levels of employee engagement, satisfaction, and com- another unique selling point of the JD-R model.

mitment have increased; (2) with a few exceptions, levels of As is exemplified by the case above, the JD-R model

job demands and job resources remained rather stable; (3) provides a common language among members of the orga-

although no firm conclusions can be drawn about the effec- nization that facilitates communication about ‘work and

tiveness of the interventions it seems that they can be linked well-being’. For many organizational members, including

to increases in some job resources. management, this is rather unknown and slippery territory

Some caveats are also worth mentioning. It might indeed for which an appropriate vocabulary is lacking. Moreover, the

be that the interventions have increased some resources, JD-R model also acts as a vehicle to understand the under-

which, in their turn, had a positive effect on employee well- lying psychological dynamics in terms of stress- and motiva-

being and commitment. But perhaps the interventions also tional processes. Finally, the JD-R framework plays a major

changed other aspects of the organization that were not role in prioritizing and implementing future actions to

assessed by the EC, such as organizational climate. Also, increase work engagement and prevent burnout. Because

perhaps more time is needed for the changes to have a the JD-R model is intuitively appealing and easy to explain,

measurable effect. For instance, employees participate only for instance by using concrete examples from employee’s or

once per year in a future oriented ‘feed forward’session with manager’s own experiences, it is well suited as a conceptual

their supervisor so that it is not very realistic to expect large, tool for understanding and guiding future actions.

short-term changes in job characteristics as a result of A logical next step is to extend the EC with an online

renewing the HR-cycle. Finally, and most importantly, eval- platform for self-management and self-enhancement that

uating a project by using a follow-up measurement is not the also builds upon the JD-R model. Based on the EC-scoring

same as investigating the effectiveness of interventions. pattern, the online system suggests for each participating

Ideally, for that a randomized control trial is needed. That employee a set of specific training modules to improve work

type of experimental design not only includes a control group engagement and prevent burnout. For instance, ‘how to

but also the randomization of employees across the inter- craft your job’, ‘be in control of your time’, ‘ how to manage

vention and control groups. Rather, the evaluation of the your boss’, ‘dealing with uncertainty’, or ‘give yourself a

current project is a final step to conclude the first cycle and break’. Following the logic of the JD-R model, these online

to provide input for the next cycle. For this particular modules aim to increase employee’s resources and reduce

project this means, for instance, that the observed stability their demands. The online platform may be used individually

should be discussed in feedback sessions with management by employees as well as collectively by teams. Currently, the

and employees. Do they recognize this? Have other things platform is being implemented and evaluated across various

perhaps changed? Did changes in one part of the organization organizations. So the next step forward is to add a perso-

cancel out those in other parts? Such questions could mark nalized, efficient online system that supports the organiza-

the start of a new cycle. tional development process toward better employee well-

being.

CONCLUSION AND OUTLOOK

Acknowledgment

In this article, it is argued and illustrated that the JD-R model

may serve as the guiding principle for an organizational This research was funded by the Research Fund KU Leuven.

130 W.B. Schaufeli

it appeared that the 9-item version of the Utrecht Work

APPENDIX A. CONSTRUCTION OF THE

Engagement Scale correlated .96 with the short 3-item

ENERGY COMPASS 4

version that is included in the EC. Moreover, both versions

As few items as possible were included in the EC in order correlated almost identically with 41 demands, resources

to minimalize the respondent’s burden, which is particular and outcomes with an average, absolute difference of only

important for survey research in organizations. In fact, .02. Finally, the internal consistencies (coefficient a) of all

23 concepts were measured with only one item, which shortened scales in the EC exceed by far the usual criterion

agrees with a recent call for the use of single-item measures of .70 with an average value of .86. In short, there is

3

in organizational research. As a rule, items from existing considerable evidence for the reliability and the validity

scales were used for the EC, but for 18 constructs self- of the EC.

formulated items had to be used because no items from In order to obtain reference data from a relevant group

existing scales were available (see Table 2). The items of the that could serve as a benchmark, the EC was administered in

5

EC were selected on face validity and correlations were a representative sample (N = 1213 ) of the Dutch working

computed with the total scale-score using data collected population aged between 18 and 65. Based on the 25th and

over the years in various samples (the total number of 75th percentiles of the frequency distributions of the scores,

respondents exceeds 50,000). A criterion of r > .80 was used cut-off values for ‘low’ and a ‘high’ scores were established

for these item-total correlations; most values were around for each construct. So when an employee scores ‘high’

.90, though. Additionally, in order to evaluate the concur- (‘low’) on a particular job demand, resource, outcome, this

rent validity, correlations with other variables were com- means his or her score is comparable with that of the 25%

pared between the short (or 1-item) scales and the original highest (lowest) scoring Dutch employees. For work engage-

longer scales. It was expected — and found — that the size of ment and burnout the cut-off for a ‘high’ score was set at

both correlations did not largely differ (i.e., <.10). For 15%. Those who score in between both cut-off values are

instance, using a total sample of over 77,000 respondents, ‘average’ compared to other Dutch workers.

4

Schaufeli, Shimazu, Hakanen, Salanova & De Witte (2016).

5

For establishing the size of the sample an error margin of 3% and

3

Fisher, Matthews & Gibson (in press). a confidence interval of 5% was used (Schaufeli, 2015a).

The JD-R model: A ‘how to’ guide 131

SELECTED BIBLIOGRAPHY

—

The Job Demands-Resources Model of Work Environment & Health, 31, 438 449; (3) the Nordic

Questionnaire for Psychosocial and Social Factors at Work

(QPS ), Elo, A. L., Skogstad, A., Dallner, M., Gamberale,

The JD-R model, which originally only included burnout, was Nordic

F. , Hottinen, V. , & Knardahl, S. (2000). User’s guide for the

first published in: Demerouti E., Bakker, A.B., Nachreiner, F. ,

QPSNordic: General Nordic Questionnaire for psychological

& Schaufeli, W.B. (2001). The Job Demands–— Resources

and social factors at work. Copenhagen: Nordic Council of

model of burnout. Journal of Applied Psychology, 86,

Ministers; (4) the Health and Safety Executive’s Management

499—512. A few year later, the extended JD-R model with

Standards Indicator Tool (HSE MSIT), Edwards, J.A. & Web-

engagement was published: Schaufeli, W.B. & Bakker, A.B.

ster, S. (2012). Psychosocial risk assessment: measurement

(2004). Job demands, job resources and their relationship

invariance of the UK Health and Safety Executive’s Manage-

with burnout and engagement: A multi-sample study. Jour-

ment Standards Indicator. Work & Stress, 26, 130—142.

nal of Organizational Behavior, 25, 293—315. Both seminal

Recently, Aacase is made in favor of using single-item mea-

papers haven been cited over 3500 times each (Google

sures by Fisher, G.G., Mattews, R.A., & Mitchell Gibbons, A.

Scholar; May, 2016), which illustrates the popularity of

(in press). Developing and investigating the use of single-

the JD-R model among researchers. Still a few year later,

item measures in organizational research. Journal of Occu-

personal resources were added to the model: Xanthopoulou,

pational Health Psychology. A first validity study oft he

D., Bakker, A.B., Demerouti, E., & Schaufeli, W.B. (2009).

Engergy Compass has been published in Dutch: Schaufeli,

Reciprocal relationships between job resources, personal

W.B. (2015a). Van burnout naar bevlogenheid Werk en wel-

resources and work engagement. Journal of Vocational

bevinden in Nederland [From burnout to work engagement:

Behavior, 74, 235—244. A conceptual overview can be found

Work and well-being in the Netherlands]. M&O, 69, 15—31.

in: Bakker, A. & Demerouti, E. (2007), The Job Demands-

The Energy Compass has also been used in: Schaufeli, W.B.

Resources (JD-R) model: State of the art. Journal of Manage-

(2015b). Engaging leadership in the Job Demands-Resources

rial Psychology, 22, 309—328. An overview of recent empiri-

Model. Career Development International, 20, 446—463.

cal findings is presented in: Schaufeli, W.B. & Taris, T. W.

(2014). A critical review of the Job Demands-Resources

Model: Implications for improving work and health. In G. Intervention (Case Example)

Bauer & O. Hämmig (Eds), Bridging occupational, organiza-

—

tional and public health (pp. 43 68). Dordrecht: Springer. For an overview (of research) on work engagement see:

The most recent review is: Taris, T. W. & Schaufeli, W.B. (in Schaufeli, W.B. (2014). What is engagement? In C. Truss,

press). The Job Demands-Resources model, its bases, appli- R. Delbridge, K. Alfes, A. Shantz, & E. Soane. (Eds.).

cations and range. In S. Clarke, T. Probst, F. Guldenmund and Employee engagement in theory and practice (pp. 15—

J. Passmore (Eds), The Wiley-Blackwell handbook of the 35). London: Routledge. For a review of case studies on

psychology of occupational safety and workplace increasing work engagement in organizations see: MacLeod,

health. Chichester: Wiley-Blackwell. D. & Clarke, N. (2009). Engaging for success: Enhancing

performance through employee engagement. Office of Pub-

Assessment (Energy Compass) lic Sector Information. London, UK: http://dera.ioe.ac.uk/

1810/1/file52215.pdf. The so-called regulative cycle has

been described by Van Strien (1977). Towards a methodology

Various questionnaires have been published to assess psy-

of psychological practice: The regulative cycle. Theory &

chosocial work factors and employee well being, for

Psychology, 7, 683—700. The relevance of engagement and

instance: (1) the Questionnaire on the Experience and Eva-

burnout for the hospitality industry has been documented in;

luation of Work (QEEW), Van Veldhoven, M., De Jonge, J.,

Salanova, M., Agut, S., & Peiró, J. M. (2005), Linking orga-

Broersen, S., Kompier, M., & Meijman, T. (2002). Specific

nizational resources and work engagement to employee

relations between psychosocial job conditions and job-

performance and customer loyalty: the mediation of service

related stress: A three-level analytic approach. Work &

climate, Journal of Applied Psychology, 90, 1217—1227 and

Stress, 16, 207—228; (2) the Copenhagen Psychosocial Ques-

Pienaar, J. & Willemse, S.A. (2008), Burnout, engagement,

tionnaire (COPSOC), Kristensen, T. , Hannertz, H., Hogh, A.,

coping and general health of service employees in the

& Borg, V. (2005). The Copenhagen Psychosocial Question-

hospitality industry, Tourism Management, 29, 1053—1063, naire (COPSOQ)–— A tool for the assessment and improvement

respectively.

of the psychosocial work environment. Scandinavian Journal

Wilmar B. Schaufeli is distinguished research professor at Leuven University, Belgium and full professor of Work

and Organizational Psychology at Utrecht University, The Netherlands (www.wilmarschaufeli.nl). In addition, he

132 W.B. Schaufeli

is also visiting professor at Loughborough Business School, UK, and Jaume I Universitat, Castellon, Spain. He was

awarded ‘Highly Cited Researcher’ by Thomson Reuters in recognition of ranking among the top 1% of most cited

researchers in the field of psychology. Initially, his research interest was on job stress and burnout, but in the last

decade his focus shifted toward positive occupational health psychology. Dr. Schaufeli is a fellow of the European

Academy of Occupational Health Psychology, a licensed occupational health psychologist, and also works part-

time as an organizational consultant (www.3ihc.nl). (Research Unit Occupational & Organizational Psychology

and Professional Learning, KU Leuven, Belgium; Department of Psychology, Utrecht University, P.O. Box 80.140,

3508 TC Utrecht, The Netherlands. Emails: [email protected]; [email protected]).